PylStenNet

Abstract:

Artificial intelligence (AI) methods are being increasingly utilized to augment physicians’ capabilities in multiple arenas, including computer-aided diagnosis. We sought to evaluate the feasibility of creating automated detection tools for pediatric surgical conditions through deep learning models using abdominal ultrasound for pyloric stenosis. Our models displays high accuracy in multiclass segmentation of two-dimensional ultrasound imaging for pyloric stenosis, producing automated labels that allow for clear visual identification of relevant structures that can subsequently be used to classify the structures as normal or abnormal, allowing for an automated determination of a diagnosis of pyloric stenosis. 

Example Segmentation Results with PylStenNet:

TBA

Reference:

S. Deoghare et al., PylStenNet: Automated detection of pyloric stenosis in ultrasound images with deep learning. In preparation, 2025.

Back to Ultrasound Projects Main Page. Back to Research Page.